COMPUTATION OF THE COMPOSITE NOISE 607 



These impulses occur at a rate which usually is not uniform. Pro- 

 vided, however, that the general conditions do not change, the rate 

 approaches uniformity if the time interval considered is sufficiently 

 large. Some of the impulses from the different sources are super- 

 imposed upon each other while others fall in the intervals between the 

 impulses from other sources. As the amount of noise increases, two 

 general phenomena are observed: First, the loudness of the noise 

 increases due to the superposition of impulses; secondly, the noise 

 becomes steadier due to the more complete filling in of relatively silent 

 intervals (20 db or more below the average). 



Definition of Terms 



The solution of the problem of computing the total noise from its 

 component parts requires the definition of a number of terms and a 

 study of the characteristics of the implied measuring instrument. 



Definition 1 

 Each individual producer of noise is referred to as a ''noise source.'' 

 Illustrations of noise sources are: For the case of room noise — the 

 conversation of one person, the noise from a typewriter or from a fan 

 in the room. A number of sources of street noise have been men- 

 tioned above in illustrating the impulsive character of common noises. 



Definition 2 



The deflections on the measuring device {sound level meter) produced 

 by the impulses of a single source are called "source peaks." 



A peak is obtained by passing a noise impulse into a sound level 

 meter. Depending on this measuring device, the characteristics of 

 a peak differ from those of an impulse. The characteristics of the 

 sound level meter, therefore, are important in connection with this 

 computation method. Three of these, the frequency response, the 

 rule of combination of the frequency components of a complex wave, 

 and the dynamic characteristic of the indicating meter, are here con- 

 sidered in detail. These are defined in the "American Tentative 

 Standards for Sound Level Meters" approved by the American Stand- 

 ards Association ^ from which the following abstracts are made : 



1. The free field frequency response of a sound level meter, provided 

 only one response is available, shall be the 40 decibel equal 

 loudness contour modified by differences between random and 

 normal free field thresholds. Methods are given in the ASA 

 specification for correcting the reading when the microphone of 

 the sound level meter responds differently to sound waves 

 arriving with different angles of incidence. 



